1323
shown that the conformation of the enzyme at pH 10.5 is greatly different from that at n e ~ t r a l i t y . The ~ ~ protonated isoleucine amino group is thought to form a salt linkage with aspartic a~id-194,3~ thereby influencing the configuration of the active site. Thus, interactions between charged groups have an important influence upon conformation. Unfavorable charge interactions or necessity for movement of the acyl group toward the solvent in the present case could move the acyl group away from histidine-57, and explain the relatively slow first-order rates, the pH independence of the reaction, and the lack of effect of urea. Conclusions It is clear that the acyl chymotrypsin derivatives pre(32) J. McConn, G. D. Fasman, and G. P. Hess, J . Mol. B i d , 39,551 (1969). (33) P. B. Sigler, D. M. Blow, B. W. Matthews, and R . Henderson, ibid., 35, 143 (1968).
pared from I and I1 differ qualitatively in their hydrolytic behavior, and that both show striking mechanistic differences in comparison with normal ester acyl enzymes. The nucleophilic attack by a functional group in the protein, releasing p-nitrophenol and giving an inhibited enzyme, that most likely takes place with the initial acyl enzyme formed from bis(4-nitrophenyl) carbonate, cannot be detected in hydrolysis of the acyl enzyme derived from 0-(4-nitrophenylene) carbonate even though both compounds are reactive nitrophenyl esters. With the acyl enzyme from 11, the pH-independent release of 4-nitrocatechol to give an active enyzme must then result from the presence of a neighboring phenoxide ion. Thus, in both cases, the mechanism of the deacylation reaction is dependent on the chemical nature of the acyl enzyme compound. Acknowledgment. This work was supported by a research grant from the National Institutes of Health.
Kinetics and Mechanism of Decarboxylation of N-Arylcarbamates. Evidence for Kinetically Important Zwitterionic Carbamic Acid Species of Short Lifetime S . L. Johnson* and Diana L. Morrison
Contribution from the Department of Biochemistry, School of Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania 15213. Received February 2, 1971 Abstract: The rates of decarboxylation of a series of substitutedN-arylcarbamates are examined as a function of pH and buffer concentration. Specific and general acid catalysis, a water reaction, and an alkaline inhibition, which is not accountable for in terms of the ionic state of the reactants, are found. The latter reaction is explainable only in terms of an unstable zwitterionic form of the carbamic acid as a reaction intermediate. The lifetime of this intermediate is estimated to be in the range of 10-*-10-10 sec. The high pH inhibition of decarboxylation shows NCor cleavage to be rate limiting for carbamates derived from weakly basic amines and proton transfer to the nitrogen atom to be rate limiting for carbamates derived from strongly basic amines. The transition state for decarboxylation has a large amount of unimolecular character with little or no participation of nucleophilic ageais. The pK, of p-nitrophenylcarbamic acid is 4.2 and the pK, for nitrogen protonation of carbamates to form the zwitterionic species is estimated to be about -4.
nly recently has the mechanism of COz transfer reactions been studied with the view of understanding the nature of the transition state(s) involved. Early studies of Faurholt and coworkers' have demonstrated the phenomenon of carbamate formation from CO, and amines, and over a limited pH range the rate of approach to equilibrium in aqueous solutions containing amines, COz, and carbonate has been measured. More recent interest in CO, transfer reactions has centered around the nature of the enzyme-bound N-carboxybiotin I, which is involved in COz transfer from carbonate to appropriate substrates such as acetylCoA, pyruvate, or urea. The investigations of Caplow3v4 and of Caplow and Yager5 have established that car-
0
(1) C. Faurholt, J . Chim. Phys. Physicochim. B i d , 22, 1 (1925), and subsequent papers. (2) For a recent review of the pertinent literature see J. Knappe, Annu. Reo. Biochem., 39, 751 (1970).
(3) M. Caplow, J . Amer. Chem. SOC.,87, 5774 (1965). (4) M. Caplow, ibid., 90, 6795 (1968).
bamates undergo acid-catalyzed decomposition and that carbamates derived from amines with pKa values less than 5 are dependent upon the pKa of the parent amine with a /3 value of +0.77. Carbamates derived from more strongly basic amines decarboxylate at nearly the same rate with a rate constant of -lo8 M-l sec-l. In addition, general acid catalysis was found for the decarboxylation of the biotin model N-carboxyimidazolidone (a = 0.9) and a striking stabilization of this carbamate was observed with metal ions such as cupric and manganous ions. In the present investigation we more fully investigate the pH dependence of carbamate decomposition over a very wide pH range, and establish free-energy relationships with a series of compounds more closely allied to each other in structure than those used by Cap10w.~ For this purpose N-arylcarbamates were used. Be( 5 ) M. Caplow and M. Yager, ibid., 89, 4513 (1967).
Johnson, Morrison / Decarboxylation of N-Arylcarbamates
1324
-4
F I
I
2
/
I
/
10
6
I
I I 14 9
PH
II
13
PH
Figure 1. Effect of pH on the rate of decarboxylation of N-(pnitropheny1)carbamate in water at 25": 0, light water; 0, heavy water.
Figure 2. Effect of pH on the rate of decarboxylation of N,N-diphenylcarbamate in water at 25".
cause these compounds undergo changes in absorbance in the ultraviolet region during decarboxylation, very small amounts of substrate could be used in the kinetic determinations, making back reactions of negligible importance. Of biological importance is the question of direct CO, transfer reactions to and from the carbon atom in carbamates. The possibility of nucleophilic reactions at the carbon atom of the carbamates used here was examined. Also of biological importance is the role of metal ions in mediating the reactions of biotin-containing enzymes. The effects of the metal ions which are effective in stabilizing N-carboxyimidazolidone5 are studied for the N-arylcarbamates, and the effects of electrophilic catalysts which are effective in bicarbonate6 and ph~sphoramidate~ decomposition are investigated here.
a kinetic experiment. A measured quantity of the stock solution, usually 10 111, is added with rapid manual mixing to 3 ml of the desired buffer in a cuvette which has been temperature equilibrated. Absorbance readings were taken at a constant wavelength in the 250-28Gnm region with a Beckman DU-2 spectrophotometer which is thermostated at 25.0 =k 0.01 '. For p-nitrophenylcarbamate, readings are taken at 380-400 nm, and, for N-cyclohexylcarbamate, readings are taken at 200-230 nm. For the latter compound and some of the arylcarbamates, a Cary-16 spectrophotometer thermostated at 25.0 i 0.03 was used to measure the absorbance changes because of the greater sensitivity of this instrument to small changes in absorbance. Plots are made of log ( A t - A,) or log ( A , - A t ) , where A , and A t are the infinity time absorbance and the absorbance at time t. A linear relationship is obtained to greater than 90% reaction; the rate constant is the slope X 2.303. Very slow reaction rates of p-nitrophenylcarbamate and diphenylcarbamate in NaOH solutions were obtained by placing a portion of the reacting solution in the Cary 16 spectrophotometer for three or more days without disturbing the cell or its contents during the time readings are taken. The end point is obtained by adding a known volume of an appropriate concentration of acetic acid to a known volume of a separate portion of the reacting mixture, and measuring the absorbance. Due to a side reaction of diphenylamine in highly basic solutions it was not possible to follow reactions of diphenylcarbamate in these media. The determination of very large rate constants was accomplished with the aid of a Durrum-Gibson stopped flow spectrophotometer. To detect buffer catalysis, dilutions were made from a stock buffer solution and KC1 was added t o maintain a constant ionic strength of 0.6 M in most cases. The pH of these buffers was measured with a Radiometer TTT-1 pH meter equipped with a Radiometer PHA630T scale expander. Values of pH higher than 13 in NaOH solutions were estimated using Harned and Heckers' activity data for NaOH.8 These values agreed within 0.02 pH unit of the measured pH values using a Radiometer combined electrode G K 2301 B, standardized against Fisher pH 11 buffer and corrected for sodium ion error with the aid of a nomograph supplied with the electrode. The pD values of heavy water solutions were calculated by adding 0.40 to the pH meter readings. A plot of kobsd os. total buffer concentration gives, in general, a straight line, the slope of which is kcat and the intercept is kink. Division of koa$by the fraction of the buffer present in the acid form gives ~ B H . To correct for pH drift in the buffer dilutions kobSd/[H +] was plotted against the buffer concentration. The intercept of this plot is k H and the slope is k,.t/[H+]. Multiplication of the slope by the average [H+] and division by the fraction of the buffer in acid form gives ~ B H .
Experimental Section Materials. Carbamates derived from primary arylamines were prepared by treating the corresponding Aldrich aromatic isocyanates (all liquids except for p-nitrophenyl isocyanate) dissolved in acetonitrile or dioxane with sodium hydroxide solution. The solid which precipitates from these solutions was dried and stored in the freezer until used. Cyclohexylammonium N-cyclohexylcarbamate from Fisher Scientific was recrystallized from benzene before use. Sodium N,N-diphenylcarbamate was prepared by stirring 19 M NaOH into a concentrated dioxane solution of N,N-diphenylcarbamoyl chloride (Fisher Scientific). The precipitate was washed with dilute NaOH and dried. The infrared spectrum of this solid shows the complete absence of the carbonyl stretching frequency of the diphenylcarbamoyl chloride starting material at 1720 cm-I. The carbamate has a low intensity stretching frequency at 1666 cm-'. During the hydrolysis of the aromatic carbamates, the Amax of the in the uv region characteristic of carbamate was replaced by, , A the aromatic amine. In the pH range 10-13 the following spectral changes were observed: p-methoxyphenylcarbamate, a band at 237 nm was replaced by a band at 292 nm; o-ethoxyphenylcarbamate, 237 nm replaced by 281 nm; phenylcarbamate, 238 nm replaced by 275 nm; p-chlorophenylcarbamate, 244 nm replaced by 290 nm; o-chlorophenylcarbamate, 248 nm replaced by 280 nm; p-nitrophenylcarbamate, 349 nm replaced by 381 nm; m-nitrophenylcarbamate, 278 nm replaced by decreased absorbance; diphenylcarbamate, 245 nm replaced by 279 nm. Only decrease in end absorption in the 230-nm region was observed for N-cyclohexylcarbamate. All buffer components are commercial products of reagent grade, with the exception of imidazole, which was recrystallized before use, and N-methylimidazole from K and K Laboratories. Deuterium oxide, 99.7 is from General Dynamics Corporation, Liquid Carbonic Division. Kinetic Determination. Stock solutions of the N-arylcarbamates were prepared in 0.01 M NaOH or methanolic 0.01 M NaOH before
z,
(6) M. M. Sharma and P.V. Dankwerfs, Trans. Faraday Soc., 59,386 (1965). (7) W. P. Jencks and M. Gilchrist, J . Amer. Chem. Soc., 86, 1410 ( 1964).
Journal of the American Chemical Society j 94:4
Resu1ts The decarboxylations of all the N-arylcarbamates examined here involve buffer catalysis. With the exception of certain amine buffers the catalysis is of the general acid type. Extensive buffer studies were carried out for N,N-diphenylcarbamate and N-(p-nitropheny1)carbamate, the results for which are shown in Tables I and 11. Ionic strength effects and possible cosolvent effects of buffer components were checked by adding KC1 and dioxane to Tris buffers containing the car(8) H. S. Harned and J. C.Hecker, ibid.,55, 4838 (1933).
/ February 23, 1972
1325 Table I. Kinetic Parameters for the Decomposition of N-(p-Nitropheny1)carbamate in Water at 25” Buffer
pKaa
p,
M
Stock buffer composition in M basic form/M basic formb
kcst,
pH rangei
M-l min-’
km, M-1 min-1
~~
Borate Carbonate Phosphate Phosphate Imidazole Imidazole Triethanolamine Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1) aminomethane in DzO Dithiothreitol Glycylglycine Glycylglycine Glycylglycine Glycylglycine Butylamine Butylamine Acet y lacetone Acetylacetone Copper( 11) acetylacetone
9.24k 10.33” 7.209 7.20 6.990 6.99 7.76~ 8.078
0.6 0.6 0.6 0.6 0.1 0.6 0.6 0.6
O.lO/O. 10 0.040/0.40 0.16/0.016 0.20/0.005 1.oo/o. 10 1 .oo/o. 10 0.80/0.40 0.70/0.30
9.15-9.08 8.84-8.91 7 . 52c 8.33-8.22 8.12-8.17 8.27-8.23 8.31-8.35 8.69-8.63
2.52 0.500 0.473 -0.077d 0 f 0.001
8.07’
0.6
1.2/0.06
8.68-8.74
0 i 0.001
0 i 0.004
0.6
0. lO/O. 10
9.97-9.93i
0.055
0.11
0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6 0.6
0.05/0.50
8.20-8.18 8.54-8.41 8.78-8.72 9.12-9.04 9.42-9.32 9.60-9.70 10.01-10.08 9.12-9.45 8.60-8.43 8.9
Arsenite Arsenite Copper(I1) glycylglycine
9.221
1.0
9.2’
0.6
9.10~ 8.2Y
10.611 10.61 8.9P 8.95m
Bromine Formaldehyde
13.29f
0.4
Isobutyraldehy de
13.9
0.6
Carbonic anhydraseh Phosphite Chloral
6.58‘ 10.04f
0.25
0.60/0.60
0.40/0.20 0.50/1.10 0.56/0.07 0.06/0.60 0.16/0.60 0.60/0.60 0.30/0.90 0.05 M CuSOa in 0 . 1 M acetylacetone buffer 1 M sodium arsenite 1 M sodium arsenite 0.05 M CuSOain 0.1 M glycylglycine buffer 0 . 1 M B r z i n 0 . 1 5M Na2C03-0.15 M NaHC03 0.1 M i n 0 . 0 5 MNa2C034).25 M NaHC03 0.11 M isobutyraldehyde in 0.15 M Na2CO3,0.15 M NaHC03 Veronal, 0.03 M 0.1 M sodium phosphite 0.05 Mchloralin0.15 M Na2C03-0. 15 M NaHCOa
0.128 0.190
-0. 246d 0.673 0.396 0.181 0.0900 -0. 0066d -0. 0019d 0.145 0.25 0 i 0.02
9.86 8.94 8.21
0.033 0.080 0 i 0.02
8.91
0 i 0.02
9.04
0 i 0.02
9.47 9.12 8.62 9.11-9.53
0.257 0.209 87 106 5.55
5.17 -0.23d 0 i 0.003
-0.261d 1.35 1.19 1.09 0.810 -0. 0073d -0. 0024d 0.29 0.33 0.33 0.16 0.15
-0.13 1.5
160
-0.008 d~ 0.003
a pK, values are for Lero ionic strength and 25” unless otherwise noted. The stock buffer and at least four dilutions are used for the determination of the catalytic coefficients. c One point only. d Linear buffer inhibition. 6 Estimated by potentiometric titration. f Aldehyde hydrate ionization: R. P. Bell and D. P. Onwood, Trans. Faraday Soc., 58,1557 (1962). 0 Estimated f r o a pK. of acetaldehyde hydrate ionization; footnotef. Worthington carbonic anhydrase ( 1 mg) in 100 ml of 0.03 M veronal buffer. The pH values are listed according to the stock and the most dilute buffer, usually 0.1 or 0.2 dilution. 7 pD = pH meter reading 0.4. B. B. Owen, J. Amer. Chem. SOC.,56, 1695 (1934). J. Bjerrum, G. Schwarzenbach, and L. G. Sillen, “Stability Constants of Metal-Ion Complexes, Part 11, Inorganic Ligands,” Chemical Society, London, 1958. R. G. Pearson and R. L. Dillon, J. Amer. Chem. Soc., 75, 2439 (1953). J. T. Edsall and J. Wyman, “Biophysical Chemistry,” Academic Press, New York, N. Y., 1958. 0 S. P. Datta and A. K. Grzyzbowski, J . Chem. Soc., 136 (1966). p R. G. Bates and G. F. Allen, J. Res. Nut. Bur. Stand., Sect. A , 64,343 (1960). * J. J. Christensen and R. M. Izatt, J . Phys. Chem., 66, 1030 (1962). E. R. B. Smith and P. K. Smith, J. Biol. Chem., 146, 187 (1942). a S. P. Datta, A. K. Grzyzbowski, and B. A. Weston, J . Chem. Soc., 792 (1963). 1 E. Breslow in “The Biochemistry of Copper,” Academic Press, New York, N. Y., 1965, p 149. u H. C. Brown, D. H. McDaniel, and 0. Haffinger in “Determination of Organic Structures by Physical Methods,” E. A. Braude and F. C. Nachod, Ed., Academic Press, New York, N. Y., 1955.
+
bamates. For diphenylcarbamate in a 0.2 M Tris buffer at pH 8.2 an increase in the ionic strength from 0.1 to 1.75 M with KC1 results in a linear rate decrease with a slope of -0.6 M-’ min-I. The effect of the addition of 1 M dioxane in the decarboxylation of p nitrophenylcarbamate in a Tris buffer is to increase the rate by 16%. These results are shown in Table 111. Tris buffer was chosen to demonstrate the solvent effects because the Tris buffer components themselves show very little catalytic activity. The pH drift in the buffers due to the added components is corrected for in these buffers by computing the quantity k/[H+] which in this case represents the catalytic coefficient for the hydronium ion. Logarithmic plots of the buffer-independent decarboxylation rate constants us. pH shown in Figures 1 and 2 indicate that only a hydrogen ion dependent term is important in the pH range 8-13 for diphenylcarba-
mate and from 5 to 12 for p-nitrophenylcarbamate. The decarboxylation rate of the latter carbamate which was measured below pH 7 becomes pH independent below pH 4 and above pH 13. The rate constant fits the general equation
where k , is the high pH-independent rate constant, ko is the low pH-independent rate constant, and K, is an apparent ionization constant. The rate levels off with decreasing pH by conversion of the carbamate to a reactive acid form, the apparent K, of which is 10-4.16M . The solid line in Figure 2 is calculated from eq 2; the solid line in Figure 1 is calculated from eq 3. On the kobsd = 107.66[H+] min-’ (2)
Johnson, Morrison / Decarboxylation of N-Arylcarbamates
1326 Table 11. Kinetic Parameters for the Decarboxylation of N,N-Diphenylcarbamate in Water at 25O Stock buffer composition,b Buffer
pK.
Ethylenediamine
p , M M basic form/M acidic form
9.91. 0.59
Glycine Glycine Glycine N-Methylimidazole Butylamine Butylamine Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Tris(hydroxymethy1)aminomethane Triethanolamine Ammonia Carbonate Carbonate Carbonate Phosphate Imidazole Hydroxylamine
pH range
kost,M-1 min-1
BE, M-l min-1
0.59/0.59 (half neutralized 8.72-8.90 1.28 M ethylenediamine) O.l0/0.90 8.85-8.84 0.20/0.80 9.20-9.21 0.30/0.70 9.41-9.46 1 .70/0.170 8.18-8.28 0.40/0.60 10.45-10.60 0.065/0.60 9.71-9.76 0.219/0.281 8.13-8.23
0.0326
0.0652
0.0937 0.0514 0.0358 3.1 i 0.09 0.00156 =k O.OOO1 0 f 0.00003 0.080
0.104 0.0643 0.0512 34 i 10 0.0026 =!= O.ooO2 O.oo00 f 0.0003 0.14
8 .07d 0.60
0.451/0.0482
9.34-9.31
0.012
0.12
8 . 07d 0.60
0.67/0.34
8.53-8.57
0.0057
0.017
8.07d 0.08
0.72/0.08
9.17-9.16
0.024
0.24
8.07d 0.20
0.40/0.20
8.66-8.51
8.07d 0.32
0.32/0.32
8.27-8.31
0
8 .07d 0.14
0,70/0.14
8.97-8.92
0.022
0.40/0.20 0.50/1.50 0.12/0.60 0.375/0.375 0.075/0.75 1.00Na2HPO4/0.010 NaHzPOa 1,00/0.100 m
8.42-8.46 9.46-9.57 9.26-9.37 9.18-9.28 9.06-9.22 7.45-7.42
0 f 0.02 0.059 0.026 0.035 0.0087 2.56
0 f 0.03 0.079 0.079 0.074 0.096 28.3
8.174-8.145 8.3-7.7
1.9 0.00 zk 0.001
21 0.00 zk 0.001
9.7W 0.60 9.78 0.60 9.78 0.60 6.95” 0.60 10.61k 0.60 10.61k 0.60 8.07d 0.60
7.76O 9.25a 10.33’ 10.331 10.33’ 7.200
0.60 0.20 0.975 0.488 0.915 3.1
6.99h 0.08 5.97l 0.30
-0.067 f
-0.20 0.001
0.00 i 0.001 0.13
a pK, values are for zero ionic strength and 25” unless otherwise noted. * The stock buffer a n d a t least four dilutions are used for the determination of the catalytic coefficients. c J. A. Partridge, J. J. Christensen, and R. M. Izatt, J. Amer. Chem. Soc., 88, 1649 (1966). Table I, footnote s. e Table I, footnote p . f Table I, footnote n. 0 Table I, footnote q. Table I, footnote O . D. D. Perrin, “Dissociation Constants of Organic Bases in Aqueous Solutions,” Butterworths, London, 1965. i E. J. King, J . Amer. Chem. SOC.,73, 155 (1951). H. C . Brown, D. H. McDaniel, and 0.Haffinger in “Determination of Organic Structures by Physical Methods,” E. A. Braude and F. C. Nachod, 0.2 M hydroxylamine added to Ed., Academic Press, New York, N. Y . ,1955, p 507. 1 H. K. Hall, J. Amer. Chem. SOC.,79,5441 (1957). Tris buffers.
Table 111. Ionic Strength and Solvent Effects in the Decarboxylation of N-(pNitropheny1)carbamate and N,N-Diphenylcarbamate at 25”
p, W
Buffer 0 . 2 MTris 0 . 2 MTris 0 . 2 MTris 0 . 2 M Tris
k , min-’
p-Nitropheny lcarbamate 0.10 8.143 0.502 0.50 8.188 0.389 0.75 8.224 0,352 1.0 8.310 0.333
0 . 2 MTris 0 . 2 M Tris 0 . 2 MTris 0 . 2 MTris 0 . 2 MTris 0 . 2 MTris 0 . 2 MTris 0.88 MTris 0.88 M Tris 0.88 MTris 0.512 Mdioxane 0.88 M Tris 0.512 Mdioxane 0.88 M Tris 0.512 Mdioxane a
pH
Diphenylcarbamate 0.10 8.141 0.50 8.286 0.75 8.278 1.0 8.287 1.5 8.329 1.75 8,350 1.75 8.370 0.08 9.221 0.08 9.209 0.08 9.217
k/[H+l X 10-7, ~ min-1 6.98 6.00 5.89 6.80
0.315 0.219 0.204 0.189 0.156 0.148 0.140 0.0467 0.0429 0.0495
4.36 4.23 3.87 3.66 3.32 3.32 3.29 7.77 6.94 8.19
0.08
9.222
0.0517
8.61
0.08
9.208
0.0569
9.19
Adjusted with KCl.
hydronium ion catalyzed portion of the pH-rate profile for p-nitrophenylcarbamate the solvent deuterium isoJournal of the American Chemical Society
/
94:4
1
tope effect kHHzO/kDDzO is 1.01; on the low pH-independent portion of the pH-rate profile the solvent deuterium isotope effect for ko is 3.3. kH refers to the hydronium ion catalytic coefficient which, according to eq 1, is given by ko/K,. The ratio of solvent deuterium isotope effects for ko/kHgives the isotope effect for K, which is 3.3. Buffer Catalysis. Diverse types of buffers were used with diphenylcarbamate and p-nitrophenylcarbamate, including Tris, phosphate, carbonate, acetylacetone, glycylglycine, glycine, dithiothreitol, imidazole, Nmethylimidazole, triethanolamine, ammonia, and butylamine. In the Brpnsted plots of this data in Figure 3 it can be seen that a good correlation is obtained with all the acid catalysts including hydronium ion, with the exception of primary ammonium ions and dithiothreitol (Table I) which exhibit large negative deviations or are inhibitory. In contrast to the results of Caplow and Y ager5 with N-carboxyimidazolidone, imidazolium ion is an effective catalyst, as is also N-methylimidazolium ion. Brpnsted QI values of 0.74, 0.76, and 0.78 are obtained for diphenylcarbamate, phenylcarbamate, and p-nitrophenylcarbamate, respectively, as shown in Figure 3. Certain catalysts which are very effective in COz hydration-dehydration were also tested. These include arsenite, phosphite, bromine, carbonic anhydrase, copper(I1) gly~ylglycine,~”and several alde(9) (a) E. Breslow in “Biochemistry of Copper,” J. Peisach, P. Aisen, and W. E. Blumberg, Ed., Academic Press, New York, N. Y . , 1965, P
February 23, 1972
1327 Table IV. Rate of Decarboxylation of Ringsubstituted N-Phenylcarbamates and of Cyclohexylamine Carbamate in Water at 2 5 O Carbamate Diuhenylp-Nitropheny lm-NitroDhenvlo-Chlorophenylm-Chloropheny1p-Chloropheny lo-Ethoxypheny1o-Methoxy pheny 1Phenylp-Methoxypheny lCyclohexylamine
pKa of leaving amine" 0.9j 1.2 2.45 3.32 3.32 3.81 4.47 4.49 4.58 5.29
kBHl
Bufferd h
h
g
g
0.05 M NaHC03/0.025 M Na2C03 0.10 M Na3P04/0.05 M Na2HP04
0.643 0.085 0.785 1.09 4.83 0.088
0.10 M NaHC03/0.02 M Na2C03 0.10 M NaHC03/0.02M Na2C03 0.19 MNa2C03/0.019M N a H C 0 3 0.05 M Na3P04/0.05 M Na2HP04 .f 0.05 M Na3P04/0.05 M Na2HP04 NaOH bufferse 0.20 M Na3P04/0.15 M Na2HPO4 p = 1.8 0.60 M Na2CO3/0.06 M NaHC03 p = 3.7
10.64
y-'
10-"kH, M-' min-'
min-
0.131 -0.0015 f 0.0003 0 f 0.005
ab
0.0045 0.0078 0.0396
0.16 0.71 0.12
0.117 0.175 0.93 2.65 0.36 0.735 1 . 2 z!z 0 . 2
0.76 0.76 0.16
ac
0.17 0.85
0.73 0.80
1.2 0.92
>0.9
0 From H. A. Sober, Ed., "Chemical Rubber Handbook of Biochemistry," Chemical Rubber Publishing Co., Cleveland, Ohio, 1968. Calculated from the bicarbonate and hydronium ion terms. c Calculated from the dianionic phosphate and hydronium ion terms. Unless otherwise noted, the ionic strength is 0.6 M. The stock buffer listed here and four dilutions were used to determine the kinetic parameters. e NaOH, 0.02-0.5 M was used. The pH before and after reaction was determined and the pH drift (less than 0.2 unit) was taken into account in calculating k=. f See Table V. 0 See Table I. h See Table 11. N. F. Hall, J. Amer. Chem. Soc., 52, 5115 (1930).
b
hydesqs No extraordinary catalysis with these agents was found. Phosphite and arsenite fit the Brgnsted line for acids of their pK, values. Acetylacetone is plotted both as its carbon acid (pK, 8.92) and as corrected for its enol form ( ~ K 6 . 2 5 ) ~ ~ ~ Amine Buffers. Of the monofunctional amine buffers, Tris and butylamine were studied in detail. With
diphenylcarbamate and p-nitrophenylcarbamate the catalytic coefficient increases with increasing percentage of the basic form of the buffer, and at low pH the
BUN$
M
BuNH2, M
Figure 4. Effect of butylamine and butylammonium ion on the rate of decarboxylation of N-(p4trophenyl)carbamate in water at 25", p = 0.6 M (KCl): (a) butylamine held constant at 0.06 M and butylammonium ion varied; (b) butylammonium ion held constant at 0.03 M and butylamine varied.
' I
L " " " " " '
Pk, PK,
Figure 3. Brqnsted plots for the general acid catalytic coefficients for the decarboxylation of N-phenylcarbamate, N-(p-nitropheny1)carbamate, and N,N-diphenylcarbamate in water at 25": Pst, phosphite; Pi, phosphate; Im, imidazole; AcAc, acetylacetone; AcAcOH, enol form of acetylacetone; Ast, arsenite; Gly, glycine; GlyGly, glycylglycine; Trn, triethanolamine. 149; (b) M. Eigen and W. Kruse quoted in Progr. Reacr. Kine?., 2, 303 (1964).
Figure 5 . Effect of substituents on the rate of hydronium ior catalyzed decarboxylation of N-arylcarbamates in water at 25" Log k H us. pK8 of leaving amine group.
buffers have either no catalytic effect or are inhibitory. Rate constants obtained in highly alkaline solutions containing the amine rule out a significant amine reaction. Plots of the rate data according to eq 4, where Johmon, Morrison
/ Decarboxylation of N-Arylcarbamates
1328 Table V. Rate of Decarboxylation of N-Phenyl- and N-o-Ethoxyphenylcarbamate at 25" Buffer6
PHd
10aki,t, min-1
N-o-Ethoxvvhenvlcarbamate 11.8'8 10.5 12.46 4.47 12.68 3.30 13.03 2.42 13.16 2.02 13.28 1.41 13.33 1.10 13.39 1.01 14.25" 0.227 33 11.43 105 10.84
0.01 M NaOH 0.04M NaOH 0.07 M NaOH 0.2 MNaOH 0.3 MNaOH 0.5 M NaOH 0.7 M NaOH 1 .O M NaOH 2.5 M NaOH 0.05 M NaaP04/0.05 M Na2HP04 0.19 M NazCOa/O.019 M NaHC03
N-Phenylcarbamate 11.30 11.81 11.89 11.95 12.23 12.63 12.82 12.91 13.3 13.4 13.4 13.5 13.5 13.6 13.6 13.7 10.47
0.002 M NaOH 0.0040M NaOH 0.0080 M NaOH 0.010 M NaOH 0.030 M NaOH 0.060 M N a O H 0.090 M NaOH 0.10 M NaOH* 0.20 M NaOHd 0.30 M NaOH* 0.40 M NaOHe 0.50M NaOHe 0.60 M NaOH" 0.70 M NaOHe 0.80 M NaOHe 0.90 M NaOHd 0140M/O .04 M NazCO3/NaHCOa
17.3 10.7 6.83 5.46 2.26 1.59 1.13 1.33 1.05 0.746 0.900 0.614 0.550 0.492 0.444 0.436 226
~ B H M-1 ,
min-1
0 i 0.001 4.83
1.02
p = 1.2
0.10 M/O .05 M NaaPOJNaHP04 p = 0.75 0.408 M/O .60 M B u N H ~ / B u N H ~ + 0.18 M/0.018 M NazC03/NaHCOa 0.15 M/O. 03 M NazCOs/NaHCO3 0.15 M/0.15 M NazCOa/NaHCOs in DzO 0.15 M/0.075 M NazCOa/NaHCOa in DzO 0.15 Mi0.05 M NazCOa/NaHCOs in DzO 0.15 M/0.05 M NapCOr/NaHCOs in DzO 0.01 M NaOD 0.03 M NaOD 0.05 M NaOD 0.07 M NaOD 0.09 M NaOD 0 . 1 M NaOD 0.3 M N a O D 0 . 6 M NaOD 0.8 MNaOD 0.8 MNaOD 1 .O M NaOD 0.10 M NaOH O.lOMNaOHand0.20 MKCl O.lOMNaOHand0.40MKCl 0.10 M N a O H a n d 0 . 6 0 M K C l
11.53
8.4
0.108 f 0.007
10.61 10.72 10.53 10.39
9.75 60.8 106 137
0.035 i 0.002 3.69 f 0.01 2.45 f 0 . 2 2.42 i 0.15
10.67
71.1
1.76
10.95
40.6
1.42 i 0.12
11.09
27.0
1.42 i 0.12
+ 0.08
0.781 0.462 0.386 0.355 0.345 0.315 0.238 0.207 0.175 0.217 0.116 1.20 1.20 1.18 1.20
12.82 13.27 13.48 13.62 13.72 13.75 14.20 14.48 14.59 14.59 14.69 12.79 12.81 12.80 12.81
Values for pH > 13 were calculated from Harned and Heckers' activity data for NaOH,*and directly measured as explained in the Experimental Section. Ionic strength 0.6 M (KC1) unless otherwise designated. The buffer composition and four dilutions were used to deH kin$. In unbuffered systems, kin$refers to kobad. In buffered systems, kint refers to the intercept values of plots of kobad termine ~ B and US. buffer concentration. Values of pH in DzO refer to the pH meter reading + O M . e No ionic strength control. kobsd
- kHIHfl
=
kB[B]
+
+
kBH[BHl kBHBIBIIBHl (4) either the amine or the ammonium ion of the buffer is kept constant while the other component varies from 0 to 0.7 M , give consistent results for dihenylcarbamate decomposition in butylamine buffers, where kB, kBH, and kBHBare 1.2 X 10-3 M-' min-l, 1.3 X 10-3 M-1 min-l, and 2.8-3.0 X 10-3 M - 2 min-l. The value of Journal of the American Chemical Society
/
94:4
k B H is negative for p-nitrophenylcarbamate in butylamine buffers and the kB value depends upon the buffer composition as shown in Figure 4. In Tris buffers the kB and kBHvalues are negligible. Substituent Effects. The effect of substituents in the aromatic ring portion of A'-arylcarbamates on the value of k H is shown in Figure 5, where log k H us. the pK, of the leaving arylamine is plotted. A linear relation is found for the arylamines with a p value of 0.71. The
February 23, 1972
1329
13
14
PH
Figure 6. Effect of pH on the rate of decarboxylation of N-(0ethoxypheny1)carbamate as a function of pH at 25" in water. Solid line calculated from eq 5 with k H = 0.82 X 1010 M-l min-1, k , = 1.8 X m i x 1 , and Ki = 6 X M . The dotted line is the rate expected if only the kH term is present.
kH value for N-carboxyimidazolidone from the data of Caplow and Yager5 and the kH value for cyclohexylamine carbamate are included in Figure 5 for comparison. Actually, the pKa value used for imidazolidone of -0.97 is only the upper limit of the pK, of the nitrogen atom of the imidazolidone molecule because the compound behaves as an 0 acid rather than as a N acid. lo All of the arylcarbamates studied are subject to buffercatalyzed decarboxylation. The buffer catalysis is fairly constant as the basicity of the leaving amine increases, as shown in Table IV. For example, using the CY value calculated from kH and the bicarbonate catalytic coefficient, CY is 0.71, 0.72, 0.76, 0.76, 0.73 in the series p-nitro, m-nitro, p-chloro, m-chloro, and H. Similarly CY calculated from kH and the biphosphate catalytic coefficient increases in the order 0.77, 0.81, 0.85, >0.9 for the series o-ethoxyl-, p-methoxyl-, o-methoxyl-, and cyclohexylamine. The large value of CY for cyclohexylamine carbamate may not be real because of the difficulty of examining catalysis in the highly basic solutions used for this substrate. Inhibition of Decarboxylation at High pH. The rate of decarboxylation of phenylcarbamate and o-ethoxyphenylcarbamate was studied in the high pH region and the results are presented in Table V and in Figures 6 and 7. At pH values greater than 13 a second hydrogen ion dependent region in the pH-rate curve is apparent. Salt effects were ruled out as the source of the high pH inhibition for phenylcarbamate because of the lack of effect of added KC1 up to 0.6 M in 0.10 M NaOH. In the case of phenylcarbamate the alkaline inhibition was previously measured by Christienson. The data in Figures 7 and 8 can be treated according to eq 5 where kH is the hydronium ion term observed over most of the pH range, kw is the water term, and Ki is the apparent alkaline inhibition constant. The kinetic parameters which best fit eq 5 are given in (5)
Table VI. For phenylcarbamate, a solvent deuterium isotope effect of 1.0 is observed for the kH term and a value of 4 is observed for k,. A very large isotope effect of ca. 15 is apparent for Ki,but these results are only approximate. (10) (1968). (11)
G.A. Olah and A . . M. White, 1.Amer. Chem.
SOC.,90, 6087
I. Christienson, Acta Chem. Scand., 18, 904 (1964).
Figure 7. Effect of pH on the rate of decarboxylation of N-phenylcarbamate at 25" in water: 0,light water; 0, heavy water; V, t h t a of Christienson; 0 , calculated using only kH in light water. ;The solid line (HzO) was calculated from eq 5 with kH = 0.36 X I 0 1 0 M-1 min-1, k , = 1.0 X 10-3 min-1, and K = 3.2 X M. 1rhe dashed line (D20) was calculated from eq 5 with kD = 0.35 X min-l, and K , = 2 X I 0lo A4-l min-l, k D z O = 0.25 X 1 bf
WEAK
IC
'\
I
I
I
/I
i
\
t W
'
\\
/
I
REACTION COORDINATE
+
,F'igure 8. Energy-reaction coordinate for the carbamate I - I t Ft. Con amine system: dotted line, weakly basic amine; s olid line, strongly basic amine.
+
Metal Ion and Special Buffer Effects. Cupric ion, 0 , O l M at pH 5.7 and 0.1 M at pH 4.2, has a negligib ile effect on the stability of p-nitrophenylcarbamate. II 1 his pH range was chosen to maximize the effect of metal ions in complexing with the carbamate in compcttition with the hydronium ion (pKa = 4.2). In addi, tion, 0.1 M cupric ion from pH 0 to 3.6 has no effect o,t I the rate of decomposition of p-nitrophenylcarbamirte. Manganous ion has a similarly negligible effect frcjm pH 1 to 4. Barium and calcium ions, 0.1-0.8 M in Tris buffers, have a small stabilizing effect on both p nit rophenylcarbamate and on diphenylcarbamate which ca 11 be accounted for by the increased ionic strength of su (:h solutions. The behavior ofp-nitrophenylcarbamate an I A diphenylcarbamate was investigated in the presence of strong carbonyl group nucleophiles. Hydroxylan 1 ine, dithiothreitol, hypochlorite, hypobromite, and hy (hoperoxide (Tables I and 11) have little effect on the dehr:arboxylation rate. With the possibility in mind thsi t metal ions might increase the rate of nucleophilic intrxaction, Ba2+,0.1-0.8 M , was added to Tris buffers COI itaining diphenylcarbamate and 0.2 M hydroxylam1 ine. Only a decrease in the rate of decarboxylation, att I ibutable to ionic strength effects, was found. Simila [sly, cupric ions added to buffer solutions of acetylacit:tone or glycylglycine (Table 11) fail to show a rate eff &.st. ~
.i(' >hnson, Morrison
Decarboxylation of N-Arylcarbamates
1330 Table VI.
Kinetic Parameters for Decarboxylation of Selected Carbamates at 25" in Water .-
pK, of leaving amine -0.97 0.9 1.2 4.62
Carbamate Imidazolinen Diphenylp-NitrophenylPhenylPhenyl- in D 2 0 o-Ethoxy pheny 1Gly cylgly cined Ammoniad Glycined N,N-Diethyldithiocarbamate. Pprrolidinedithiocarbamatee
pKcc
10'0krr, M-1 min-1
ko, imin-1
4.2
4.61 >: 103
4.47 8.13 9.25 9.6 10.93
5.33 5.25 5.18 3,371
1 . x ;< 104 3.36 ;< lo4 4.98 : K 104 6.0
3.5 x 10-5 0.45 x 0.78 x 10-2 0.36 0.35 0.82 0.28 0.60 0.75 7.76 x
11.27
3.250
0 .o;;! 1
6.6
x
k,, min-1 2 . 5 x 10-3 Kaythen the ko/kHratio (Kc in eq 11) is equal to K,. If the opposite situation holds, K, > Ka*, the ko/kHratio is equal to Ka+. With carbamates derived from weakly basic amines there is, therefore, some uncertainty as to the meaning of the apparent pK values obtained by kinetic means. However, we are convinced that Kat is much larger than K, by the following arguments: the kinetically obtained pK, value of 4.2 for p-nitrophenylcarbamate is about one unit lower than the corresponding values for carbamates derived from ammonia and aliphatic amines. This suggests that the oxygen atom in the anionic (15) In analogy with the 23 kcal/mol activation energy for the ka term of N-carboxyimidazolidone, measured by Caplow and Yager.6 (16) F. J. W. Roughton and L. Rossi-Bernardi in "COZ: Chemical, Biochemical, and Physiological Aspects," NASA Symposium SP-188, Haverford Pa., 1968, p 41.
Johnson, Morrison
Decarboxylation of N-Arylcarbamates
1332
carbamate is more strongly basic than the nitrogen atom, because larger substituent effects on the pK,, would be expected if N-protonation occurred. The pK, difference between p-nitroaniline and ammonia is; 8 units and a similarly large pKa difference would be expected in the corresponding carbamates. In fact, the pKa difference of p-nitrophenylcarbamic acid and carbamic acid is identical with that obtained in the corresponding carbon series. p-Nitrophenylacetic acid17 has a pK, value of 3.85 which is 0.9 pKa unit smaller than that for acetic acid. Furthermore, the difference in the kinetically obtained pK, values of carbamic acids and the dithiocarbamic acids of -2 pK, units1*is similar to the difference in the pKa values obtained for acetic acid us. thioacetic acid (4.8 us. 3.2) or 1.6 units.Ig In comparing the pK, value of 3.76 for carbonic acid13 with 5.25 for carbamic acid the amino group replacement for a hydroxyl group increases the basicity of the adjacent oxygen, indicating a significant resonance interaction of the nitrogen atom; inductive effects would predict the opposite result. From the above data and arguments, as well as from those arguments made by cap lo^,^ the rate-controlling step for the decarboxylation of carbamates derived from amines with pKa values greater than -5-6 is nitrogen protonation. Carbamates derived from these more basic amines decarboxylate with the same specific rate c0nstant18~of -lo8 M-1 sec-1, making it possible to calculate the basicity of the nitrogen atom of these carbamates using the known relationship between the rate of proton transfer and the pK, difference between the donor and acceptor species. l 4 Since diffusion-controlled protonation in the thermodynamically controlled direction occurs with a specific rate constant of -lolo M-I sec-I, a specific rate constant of lo*M-' sec-' for carbamate protonation means that the pKa of the carbamate is -2 units smaller than the pKa of hydronium ion which is -1.8. Therefore, the pK, of the carbamate nitrogen is estimated to be --4. The pKa of the nitrogen atom in carboxylic acid amides and carbamic acid esters is less than 0 to - 2 and - 2 to - 3, respectively, which is the pK due to oxygen protonation.20 The estimated pK of the nitrogen in carbamate ions is more in line with the pK value of --4 reportedz1 for nitrogen protonation of N,Ndiisopropylcarbamic acid. Substituent Effects. The structural effects on the carbamate stability are shown in Figure 5 , where the slope of the line drawn through the arylamine substituents is 0.71. This value is similar to that of 0.77 found by Caplow* for amine substituents with a great deal of structure variation, and therefore scatter. The linear dependence on pK, for bases of pK up to 5 and the leveling off at high pK were found by Caplow4 and are indicated here by the point for cyclohexylamine in (17) J. F. S. Dippy, J. Chem. Soc., 357 (1938). (18) Diethyldithiocarbamate has a pKe of 3.37 at 25" (K. I. Aspila, S. J. Joris, and C. L. Chakrabarti, J. Phys. Chem., 74, 3625 (1970)). The pKa values of the dithiocarbamates derived from diisopropylamine, pyrrollidine, dimethylamine, dibutylamine, and dipropylamine are 4.2, 3.25 (S. J. Joris, K . I. Aspila, and C. L. Chakrabarti, Anal. Chem., 41, 1441 (1969)); 3.2, 3.9, and 3.6, respectively ( K . I . Aspila, V. S. Sastri, and C. L. Chakrabarti, Talanta, 16, 1099 (1969)). (19) R. Barnett and W. P. Jencks, J . Amer. Chem. SOC.,89, 5963 (1967); R. Barnett and W. P. Jencks, ibid., 91, 6758 (1969). (20) E. M. Arnett, Progr. Phys. Org. Chem., 1, 233 (1963); V. C. Armstrong and R. B. Moodie, J. Chem. Soc., 275 (1968). (21) V. C. Armstrong, C. W. Farlow, and R . B. Moodie, Chem. Commun., 1362 (1968).
Journal of the American Chemical Society
1 94:4 /
Figure 5. This break is indicative of a change in ratedetermining step in the decarboxylation reaction. The energy profile in Figure 8 depicts the deductions made above. The ground-state energy levels in both reactants and products are lowered for weakly basic amines compared with strongly basic amines, and the reverse situation is shown for the zwitterionic intermediate. A great deal of the substituent effect in kH for weakly basic amines comes from the basicity difference since kH is equal to kb/Ka+,and is reflected in the K,. values, which should parallel in substituent effect the K, values of the parent amine, giving a p value near unity. For weakly basic amines shown in the dotted line, the right-hand barrier for C-N cleavage has to be the highest barrier in the energy profile. Strongly basic amines will have a somewhat lower energy for the zwitterion form than the weakly basic amine and must have the second barrier for C-N cleavage lower than the left-hand barrier for proton transfer. This means that the substitutent effect for the C-N bond cleavage process is greater than for N-protonation. This behavior is to be expected because C 0 2 ,in analogy with acyl groups which have a 0value of 1.622compared with a value of 1.0 for the proton, should be more electropositive than a proton, and therefore the central N atom of the zwitterionic carbamic acid should be more sensitive to the removal of CO, than of the proton. Buffer Catalysis. The buffer-catalyzed term for carbamates derived from strong bases is explainable in terms of transition-state V in which rate-limiting protonation to form the zwitterionic carbamate takes place. The C-N bond remains intact because of the kinetic demonstration of an intermediate. Buffer catalysis of carbamates derived from weakly basic amines is described by transition state VI in which N-C bond breaking is concerted with the nitrogen protonation. Closely analogous transition states, VI1 for strongly basic N and VI11 for weakly basic N, for the general dS
V, pK
